##// END OF EJS Templates
rust-discovery: using from Python code...
rust-discovery: using from Python code As previously done in other topics, the Rust version is used if it's been built. The version fully in Rust of the partialdiscovery class has the performance advantage over the Python version (actually using the Rust MissingAncestor) if the undecided set is big enough. Otherwise no sampling occurs, and the discovery is reasonably fast anyway. Note: it's hard to predict the size of the initial undecided set, it can depend on the kind of topological changes between the local and remote graphs. The point of the Rust version is to make the bad cases acceptable. More specifically, the performance advantages are: - faster sampling, especially takefullsample() - much faster addmissings() in almost all cases (see commit message in grandparent of the present changeset) - no conversion cost of the undecided set at the interface between Rust and Python == Measurements with big undecided sets For an extreme example, discovery between mozilla-try and mozilla-unified (over one million undecided revisions, same case as in dbd0fcca6dfc), we get roughly a x2.5/x3 better performance: Growing sample size (5% starting with 200): time goes down from 210 to 72 seconds. Constant sample size of 200: time down from 1853 to 659 seconds. With a sample size computed from number of roots and heads of the undecided set (`respectsize` is `False`), here are perfdiscovery results: Before ! wall 9.358729 comb 9.360000 user 9.310000 sys 0.050000 (median of 50) After ! wall 3.793819 comb 3.790000 user 3.750000 sys 0.040000 (median of 50) In that later case, the sample sizes are routinely in the hundreds of thousands of revisions. While still faster, the Rust iteration in addmissings has less of an advantage than with smaller sample sizes, but one sees addcommons becoming faster, probably a consequence of not having to copy big sets back and forth. This example is not a goal in itself, but it showcases several different areas in which the process can become slow, due to different factors, and how this full Rust version can help. == Measurements with small undecided sets In cases the undecided set is small enough than no sampling occurs, the Rust version has a disadvantage at init if `targetheads` is really big (some time is lost in the translation to Rust data structures), and that is compensated by the faster `addmissings()`. On a private repository with over one million commits, we still get a minor improvement, of 6.8%: Before ! wall 0.593585 comb 0.590000 user 0.550000 sys 0.040000 (median of 50) After ! wall 0.553035 comb 0.550000 user 0.520000 sys 0.030000 (median of 50) What's interesting in that case is the first addinfo() at 180ms for Rust and 233ms for Python+C, mostly due to add_missings and the children cache computation being done in less than 0.2ms on the Rust side vs over 40ms on the Python side. The worst case we have on hand is with mozilla-try, prepared with discovery-helper.sh for 10 heads and depth 10, time goes up 2.2% on the median. In this case `targetheads` is really huge with 165842 server heads. Before ! wall 0.823884 comb 0.810000 user 0.790000 sys 0.020000 (median of 50) After ! wall 0.842607 comb 0.840000 user 0.800000 sys 0.040000 (median of 50) If that would be considered a problem, more adjustments can be made, which are prematurate at this stage: cooking special variants of methods of the inner MissingAncestors object, retrieving local heads directly from Rust to avoid the cost of conversion. Effort would probably be better spent at this point improving the surroundings if needed. Here's another data point with a smaller repository, pypy, where performance is almost identical Before ! wall 0.015121 comb 0.030000 user 0.020000 sys 0.010000 (median of 186) After ! wall 0.015009 comb 0.010000 user 0.010000 sys 0.000000 (median of 184) Differential Revision: https://phab.mercurial-scm.org/D6430

File last commit:

r42818:b53633d3 default
r42972:4d20b1fe default
Show More
state.py
220 lines | 8.3 KiB | text/x-python | PythonLexer
# state.py - writing and reading state files in Mercurial
#
# Copyright 2018 Pulkit Goyal <pulkitmgoyal@gmail.com>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
"""
This file contains class to wrap the state for commands and other
related logic.
All the data related to the command state is stored as dictionary in the object.
The class has methods using which the data can be stored to disk in a file under
.hg/ directory.
We store the data on disk in cbor, for which we use the CBOR format to encode
the data.
"""
from __future__ import absolute_import
from .i18n import _
from . import (
error,
util,
)
from .utils import (
cborutil,
)
class cmdstate(object):
"""a wrapper class to store the state of commands like `rebase`, `graft`,
`histedit`, `shelve` etc. Extensions can also use this to write state files.
All the data for the state is stored in the form of key-value pairs in a
dictionary.
The class object can write all the data to a file in .hg/ directory and
can populate the object data reading that file.
Uses cbor to serialize and deserialize data while writing and reading from
disk.
"""
def __init__(self, repo, fname):
""" repo is the repo object
fname is the file name in which data should be stored in .hg directory
"""
self._repo = repo
self.fname = fname
def read(self):
"""read the existing state file and return a dict of data stored"""
return self._read()
def save(self, version, data):
"""write all the state data stored to .hg/<filename> file
we use third-party library cbor to serialize data to write in the file.
"""
if not isinstance(version, int):
raise error.ProgrammingError("version of state file should be"
" an integer")
with self._repo.vfs(self.fname, 'wb', atomictemp=True) as fp:
fp.write('%d\n' % version)
for chunk in cborutil.streamencode(data):
fp.write(chunk)
def _read(self):
"""reads the state file and returns a dictionary which contain
data in the same format as it was before storing"""
with self._repo.vfs(self.fname, 'rb') as fp:
try:
int(fp.readline())
except ValueError:
raise error.CorruptedState("unknown version of state file"
" found")
return cborutil.decodeall(fp.read())[0]
def delete(self):
"""drop the state file if exists"""
util.unlinkpath(self._repo.vfs.join(self.fname), ignoremissing=True)
def exists(self):
"""check whether the state file exists or not"""
return self._repo.vfs.exists(self.fname)
class _statecheck(object):
"""a utility class that deals with multistep operations like graft,
histedit, bisect, update etc and check whether such commands
are in an unfinished conditition or not and return appropriate message
and hint.
It also has the ability to register and determine the states of any new
multistep operation or multistep command extension.
"""
def __init__(self, opname, fname, clearable, allowcommit, reportonly,
continueflag, stopflag, cmdmsg, cmdhint, statushint,
abortfunc, continuefunc):
self._opname = opname
self._fname = fname
self._clearable = clearable
self._allowcommit = allowcommit
self._reportonly = reportonly
self._continueflag = continueflag
self._stopflag = stopflag
self._cmdmsg = cmdmsg
self._cmdhint = cmdhint
self._statushint = statushint
self.abortfunc = abortfunc
self.continuefunc = continuefunc
def statusmsg(self):
"""returns the hint message corresponding to the command for
hg status --verbose
"""
if not self._statushint:
hint = (_('To continue: hg %s --continue\n'
'To abort: hg %s --abort') % (self._opname,
self._opname))
if self._stopflag:
hint = hint + (_('\nTo stop: hg %s --stop') %
(self._opname))
return hint
return self._statushint
def hint(self):
"""returns the hint message corresponding to an interrupted
operation
"""
if not self._cmdhint:
return (_("use 'hg %s --continue' or 'hg %s --abort'") %
(self._opname, self._opname))
return self._cmdhint
def msg(self):
"""returns the status message corresponding to the command"""
if not self._cmdmsg:
return _('%s in progress') % (self._opname)
return self._cmdmsg
def continuemsg(self):
""" returns appropriate continue message corresponding to command"""
return _('hg %s --continue') % (self._opname)
def isunfinished(self, repo):
"""determines whether a multi-step operation is in progress
or not
"""
if self._opname == 'merge':
return len(repo[None].parents()) > 1
else:
return repo.vfs.exists(self._fname)
# A list of statecheck objects for multistep operations like graft.
_unfinishedstates = []
def addunfinished(opname, fname, clearable=False, allowcommit=False,
reportonly=False, continueflag=False, stopflag=False,
cmdmsg="", cmdhint="", statushint="", abortfunc=None,
continuefunc=None):
"""this registers a new command or operation to unfinishedstates
opname is the name the command or operation
fname is the file name in which data should be stored in .hg directory.
It is None for merge command.
clearable boolean determines whether or not interrupted states can be
cleared by running `hg update -C .` which in turn deletes the
state file.
allowcommit boolean decides whether commit is allowed during interrupted
state or not.
reportonly flag is used for operations like bisect where we just
need to detect the operation using 'hg status --verbose'
continueflag is a boolean determines whether or not a command supports
`--continue` option or not.
stopflag is a boolean that determines whether or not a command supports
--stop flag
cmdmsg is used to pass a different status message in case standard
message of the format "abort: cmdname in progress" is not desired.
cmdhint is used to pass a different hint message in case standard
message of the format "To continue: hg cmdname --continue
To abort: hg cmdname --abort" is not desired.
statushint is used to pass a different status message in case standard
message of the format ('To continue: hg cmdname --continue'
'To abort: hg cmdname --abort') is not desired
abortfunc stores the function required to abort an unfinished state.
continuefunc stores the function required to finish an interrupted
operation.
"""
statecheckobj = _statecheck(opname, fname, clearable, allowcommit,
reportonly, continueflag, stopflag, cmdmsg,
cmdhint, statushint, abortfunc, continuefunc)
if opname == 'merge':
_unfinishedstates.append(statecheckobj)
else:
_unfinishedstates.insert(0, statecheckobj)
addunfinished(
'update', fname='updatestate', clearable=True,
cmdmsg=_('last update was interrupted'),
cmdhint=_("use 'hg update' to get a consistent checkout"),
statushint=_("To continue: hg update")
)
addunfinished(
'bisect', fname='bisect.state', allowcommit=True, reportonly=True,
statushint=_('To mark the changeset good: hg bisect --good\n'
'To mark the changeset bad: hg bisect --bad\n'
'To abort: hg bisect --reset\n')
)
def getrepostate(repo):
# experimental config: commands.status.skipstates
skip = set(repo.ui.configlist('commands', 'status.skipstates'))
for state in _unfinishedstates:
if state._opname in skip:
continue
if state.isunfinished(repo):
return (state._opname, state.statusmsg())